xuanyuanking commented on a change in pull request #30148:
URL: https://github.com/apache/spark/pull/30148#discussion_r519544344
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
##########
@@ -878,6 +878,14 @@ class Analyzer(
case _ => None
}
+ val isStreamingTmpView = tmpView.nonEmpty && tmpView.get.isStreaming
Review comment:
@cloud-fan @HeartSaVioR Bring back the
discussion(https://github.com/apache/spark/pull/30148#discussion_r517785184)
here:
After taking a further look, moving the checking logic in the Analyser rule
will break both SQL and DataSet API on the reading streaming temp view.
I can only block the streaming temp view reading in the DataSet side, but it
breaks the consistency between both APIs, as we don't have streaming keywords
in SQL to distinguish the table/view is whether streaming or not. So I just
unify the code path and add comments to emphasize the current behavior of
`spark.table` and `spark.read.table`.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
